随着航天器平台负载能力提升和空间信息技术的发展,有效载荷数据流蕴含的信息越来越复杂。运控管理时一般通过分析有效载荷实时数据流,判断有效载荷工作状态,进行有效载荷故障诊断。如果通过人工判读有效载荷实时数据流,工作量比较大,容易出现错误。通过分析有效载荷遥测数据流的特点,以经典的数据流分析框架CluStream算法为基础,提出一种改进算法,对有效载荷数据流进行实时处理,实现有效载荷工作状态的自动分析。仿真结果表明,改进的数据流聚类方法能够应用于有效载荷工作状态分析,且具有计算量小、实时性高和分类准确的优点。改进的数据流分析方法还可以应用到空间有效载荷数据处理。 As payload capabilities of spacecraft platforms increase and space information technologies evolve, the information contained in payload data streams becomes more and more complex. Operational management of the general analysis of real-time data stream through the payload to determine the working status of the payload, the payload fault diagnosis. If you manually interpret the payload real-time data stream, the workload is relatively large, prone to error. By analyzing the characteristics of payload telemetry data stream, an improved algorithm is proposed based on CluStream, a classic data flow analysis framework. The payload data stream is processed in real time to automatically analyze the working status of payload. The simulation results show that the improved data flow clustering method can be applied to the working state analysis of payload, and has the advantages of small computation, high real-time and accurate classification. Improved dataflow analysis methods can also be applied to spatial payload data processing.